Spaces:
Running
Running
- README.md +2 -2
- src/classifier.py +10 -4
README.md
CHANGED
@@ -62,7 +62,7 @@ python app.py
|
|
62 |
2. Open your web browser and navigate to the provided URL (typically http://localhost:7860)
|
63 |
|
64 |
3. Choose your input method:
|
65 |
-
- Upload an audio file
|
66 |
- Enter lyrics text
|
67 |
|
68 |
4. Adjust generation parameters:
|
@@ -75,7 +75,7 @@ python app.py
|
|
75 |
## Models Used
|
76 |
|
77 |
- **Genre Classification**:
|
78 |
-
- Audio: `
|
79 |
- Text: `facebook/bart-large-mnli` (Zero-shot classification)
|
80 |
- **Lyric Generation**: `gpt2-medium`
|
81 |
|
|
|
62 |
2. Open your web browser and navigate to the provided URL (typically http://localhost:7860)
|
63 |
|
64 |
3. Choose your input method:
|
65 |
+
- Upload an audio file (supports .mp3, .wav, .ogg, .flac)
|
66 |
- Enter lyrics text
|
67 |
|
68 |
4. Adjust generation parameters:
|
|
|
75 |
## Models Used
|
76 |
|
77 |
- **Genre Classification**:
|
78 |
+
- Audio: `superb/wav2vec2-base-superb-gc` (Pre-trained on music genre classification)
|
79 |
- Text: `facebook/bart-large-mnli` (Zero-shot classification)
|
80 |
- **Lyric Generation**: `gpt2-medium`
|
81 |
|
src/classifier.py
CHANGED
@@ -13,10 +13,10 @@ class MusicGenreClassifier:
|
|
13 |
model="facebook/bart-large-mnli"
|
14 |
)
|
15 |
|
16 |
-
# For audio classification, we'll use a pre-trained model
|
17 |
self.audio_classifier = pipeline(
|
18 |
"audio-classification",
|
19 |
-
model="
|
20 |
)
|
21 |
|
22 |
self.genres = [
|
@@ -29,7 +29,11 @@ class MusicGenreClassifier:
|
|
29 |
try:
|
30 |
# Load audio using librosa (handles more formats)
|
31 |
waveform, sample_rate = librosa.load(audio_path, sr=16000)
|
32 |
-
|
|
|
|
|
|
|
|
|
33 |
except Exception as e:
|
34 |
raise ValueError(f"Error processing audio file: {str(e)}")
|
35 |
|
@@ -37,8 +41,10 @@ class MusicGenreClassifier:
|
|
37 |
"""Classify genre from audio file."""
|
38 |
try:
|
39 |
waveform = self.process_audio(audio_path)
|
40 |
-
predictions = self.audio_classifier(waveform)
|
41 |
# Get the top prediction
|
|
|
|
|
42 |
top_pred = max(predictions, key=lambda x: x['score'])
|
43 |
return top_pred['label'], top_pred['score']
|
44 |
except Exception as e:
|
|
|
13 |
model="facebook/bart-large-mnli"
|
14 |
)
|
15 |
|
16 |
+
# For audio classification, we'll use a different pre-trained model
|
17 |
self.audio_classifier = pipeline(
|
18 |
"audio-classification",
|
19 |
+
model="superb/wav2vec2-base-superb-gc"
|
20 |
)
|
21 |
|
22 |
self.genres = [
|
|
|
29 |
try:
|
30 |
# Load audio using librosa (handles more formats)
|
31 |
waveform, sample_rate = librosa.load(audio_path, sr=16000)
|
32 |
+
# Convert to torch tensor and ensure proper shape
|
33 |
+
waveform = torch.from_numpy(waveform).float()
|
34 |
+
if len(waveform.shape) == 1:
|
35 |
+
waveform = waveform.unsqueeze(0)
|
36 |
+
return waveform
|
37 |
except Exception as e:
|
38 |
raise ValueError(f"Error processing audio file: {str(e)}")
|
39 |
|
|
|
41 |
"""Classify genre from audio file."""
|
42 |
try:
|
43 |
waveform = self.process_audio(audio_path)
|
44 |
+
predictions = self.audio_classifier(waveform, top_k=1)
|
45 |
# Get the top prediction
|
46 |
+
if isinstance(predictions, list):
|
47 |
+
predictions = predictions[0]
|
48 |
top_pred = max(predictions, key=lambda x: x['score'])
|
49 |
return top_pred['label'], top_pred['score']
|
50 |
except Exception as e:
|